Long-term efficacy of interferon therapy in patients with chronic hepatitis B virus infection in Japan

Transcription

1 J Gastroenterol (12) 47: DOI 1.17/s ORIGINAL ARTICLE LIVER, PANCREAS, AND BILIARY TRACT Long-term efficacy of interferon therapy in patients with chronic hepatitis B virus infection in Japan Fumitaka Suzuki Yasuji Arase Yoshiyuki Suzuki Norio Akuta Hitomi Sezaki Yuya Seko Yusuke Kawamura Tetsuya Hosaka Masahiro Kobayashi Satoshi Saito Kenji Ikeda Mariko Kobayashi Hiromitsu Kumada Received: 5 October 11 / Accepted: 5 January 12 / Published online: 24 February 12 Ó Springer 12 Abstract Background Few studies have investigated the long-term effects of interferon (IFN) therapy for chronic hepatitis B (CHB). In this retrospective study, we investigated the efficacy of and predictors of response to IFN therapy in CHB patients. Methods We analyzed data for 615 Japanese CHB patients (hepatitis B e antigen [HBeAg]-positive 414, HBeAg-negative 1) treated with IFN, and conducted follow up for a median duration of 8.1 years (range ). Responders were defined as patients who showed continuously normalized alanine transaminase (ALT) levels, HBeAg clearance, and low hepatitis B virus (HBV) DNA levels at 6 months post-treatment or for a span of more than 6 months until each test point at 1, 3, 5, and 1 years. Results The IFN response rates of all patients were 21, 18, 21, 23, and 25% at 6 months and 1, 3, 5, and 1 years, respectively. On multivariate analysis, significant determinants of the outcome of IFN therapy were as follows: at 6 months and 1 year, young age, low HBV DNA levels, and long duration of treatment; at 3 years, long duration of F. Suzuki (&) Y. Arase Y. Suzuki N. Akuta H. Sezaki Y. Seko Y. Kawamura T. Hosaka M. Kobayashi S. Saito K. Ikeda H. Kumada Department of Hepatology, Toranomon Hospital, Toranomon, Minato-ku, Tokyo , Japan fumitakas@toranomon.gr.jp M. Kobayashi Research Institute for Hepatology, Toranomon Hospital, Tokyo, Japan treatment, young age, and high level of albumin; at 5 years, high level of albumin, female, and pretreated with IFN; and at 1 years, HBeAg-negative. Sixty-nine of the 615 patients (11%) achieved seroclearance of hepatitis B surface antigen (HBsAg). On multivariate analysis, age C3 years, HBV genotype A, and male were all independent factors predicting the achievement of HBsAg seroclearance. Conclusion HBeAg, HBV DNA level, age, sex, albumin, duration of treatment, pretreatment with IFN, and HBV genotype were important factors in determining long-term response to IFN therapy. Keywords Interferon Hepatitis B virus Chronic hepatitis B Genotype Hepatitis B surface antigen Abbreviations CHB Chronic hepatitis B HBV Hepatitis B virus IFN Interferon HBeAg Hepatitis B e antigen ALT Alanine transaminase MU Million units HBsAg Hepatitis B surface antigen CLEIA Chemiluminescent enzyme immunoassay bdna Branched-chain DNA probe assay TMA-HPA Transcription-mediated amplification and hybridization protection assay PCR Polymerase chain reaction ELISA Enzyme-linked immunosorbent assay AST Aspartate transaminase AFP a Fetoprotein OR Odds ratio CI Confidence interval HCC Hepatocellular carcinoma

2 J Gastroenterol (12) 47: Introduction Hepatitis B virus (HBV) infection is a common disease that can induce a chronic carrier state and is associated with the risk of developing progressive disease and hepatocellular carcinoma [1]. Interferon (IFN) and several nucleoside/ nucleotide analogues such as lamivudine, adefovir dipivoxil, entecavir, and tenofovir disoproxil fumarate are currently approved as treatments for chronic hepatitis B (CHB) in most countries [2 5]. Successful treatment of CHB with clearance of hepatitis B e antigen (HBeAg), reduction in serum HBV DNA levels, and normalization of alanine transaminase (ALT) levels is associated with a favorable long-term outcome, independent of the antiviral drug used [6, 7]. A meta-analysis of IFN therapy published in 1993 reviewed 15 randomized controlled studies involving 837 adult patients who received IFN-a at doses of 5 1 million units (MU) administered at intervals ranging from daily to three times weekly for 4 6 months [8]. Clearance of HBeAg was noted in 33% of the treated patients compared with 12% of controls. Elimination of detectable HBV DNA and normalization of ALT levels were also more common in the treated patients than in the controls. The major pretreatment factors that correlated with a response were high ALT levels [9 11], low HBV DNA levels [9, 1], female sex, and elevated liver activity and fibrosis on liver biopsy [8]. Another recent meta-analysis of 24 randomized controlled trials concluded that the rates of persistent ALT normalization, clearance of HBeAg, and sustained elimination of HBV DNA (determined by hybridization) induced by IFN therapy were approximately 25% greater than the rates for controls. A more recent meta-analysis report showed that IFN increased the incidence of HBeAg and hepatitis B surface antigen (HBsAg) seroclearance after long-term follow up of 3 7 years [12]. However, specific data on the long-term effects of IFN therapy (median follow-up duration of 8.1 years), particularly among the Japanese, are limited. Moreover, few reports have investigated factors predicting the achievement of HBsAg seroclearance. To further evaluate factors influencing clinical outcome, we performed a retrospective cohort study on CHB patients treated with IFN in our hospital. Patients and methods Patients We retrospectively examined 615 Japanese patients (151 females and 464 males) who commenced IFN treatment between June 1984 and April 8 in the Department of Table 1 Characteristics of patients at commencement of interferon therapy Demographic data Total number 615 Sex, female/male 151/464 Age, years (range) 35 (15 68) Previously treated with interferon (%) Duration of treatment, weeks (range) 26 (4 981) Follow-up period, years (range) 8.1 ( ) Laboratory data Aspartate transaminase, IU/L (range) 72 (18 99) Alanine transaminase, IU/L (range) 138 ( ) Bilirubin, mg/dl (range).7 (.2 8.8) Albumin, g/dl (range) 3.9 ( ) Platelets, 91 3 /ll (range) 174 (48 5) Staging of liver histology (F/1/2/3/4/ND) 8/77/185/162/72/111 Serum HBV DNA, log copies/ml (range) [7.6 (\2.6 to [7.6) HBeAg (positive/negative) 414/1 HBV genotype 24/37/54/1/1/1/47 (A/B/C/D/H/B? C/unknown) Values are expressed as medians and ranges (in parentheses) or as numbers and percentages (in parentheses) HBV hepatitis B virus, HBeAg hepatitis B e antigen, ND not done Hepatology at Toranomon Hospital (Table 1). Several of the patients have been included in previous reports [13 15]. All enrolled patients were followed up for a range of years from completion of IFN treatment, with a median follow-up duration of 8.1 years. Before the commencement of IFN treatment, all patients had been positive for HBsAg in the serum for more than 6 months, and all were confirmed to have hepatitis caused by HBV and not by another vector, such as infection with hepatitis C virus or autoimmune hepatitis. None had a history of drug abuse or alcoholic hepatitis, and none had received nucleoside/ nucleotide analogue therapy. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki and was approved by the Toranomon Hospital Ethics Committee. Informed consent was obtained from each patient. Interferon therapy and assessment of response to therapy Patients received 3 12 MU of IFN-a or IFN-b (Sumiferon: Dainippon Sumitomo Pharma, Osaka, Japan; Canferon A: Takeda Chemical Industries, Osaka, Japan; Intron A: Schering-Plough MSD KK, Osaka, Japan; and Feron: Toray, Tokyo, Japan). The durations and regimens of treatment were as follows: 4 weeks (89 patients; daily for

3 816 J Gastroenterol (12) 47: weeks), 26 weeks (27 patients; daily for 4 weeks followed by 2 or 3 times a week), 52 weeks (13 patients; 2 or 3 times a week), 14 weeks (8 patients; 2 or 3 times a week), and more than 14 weeks (73 patients; 2 or 3 times a week). The median duration of treatment was 26 weeks (range 4 981). The numbers of responders were evaluated at 6 months and 1, 3, 5, and 1 years after the completion of IFN therapy. In the baseline HBeAg-positive patients, responders were defined as patients who showed normalization of serum ALT level (normal level 6 3 IU/L), HBeAg clearance, and low HBV DNA level (\5 log copies/ml) at 6 months after completion of IFN therapy. In addition, baseline HBeAg-positive patients who showed continuous normalization of ALT levels, HBeAg clearance, and low HBV DNA level for more than 6 months until each test point at 1, 3, 5, and 1 years after completion of IFN therapy were also classified as responders. In the baseline HBeAg-negative patients, responders were defined as those who showed sustained normalization of ALT level and low HBV DNA level (\4 log copies/ml) for more than 6 months until each test point after completion of IFN therapy. All patients not considered to be responders were termed non-responders. Patients receiving other therapies (IFN or nucleoside/nucleotide analogues) after the completion of IFN therapy were also termed non-responders. Blood tests and serum viral markers Routine biochemical tests were performed monthly via standard procedures during and for the first 12 months following the completion of IFN treatment and at least every 2 months thereafter. Levels of HBsAg, HBeAg, and anti-hbe were determined using radioimmunoassay kits (Abbott Diagnostics, Chicago, IL, USA) or a chemiluminescent enzyme immunoassay (CLEIA; Lumipulse System; Fujirebio, Tokyo, Japan). HBV DNA levels were measured using a branched-chain DNA probe assay (bdna) (Chiron Laboratory Service, Van Nuys, CA, USA), a transcription-mediated amplification and hybridization protection assay (TMA-HPA) (Chugai Diagnostics Science, Tokyo, Japan), or a polymerase chain reaction (PCR)-based assay (COBAS Amplicor HBV Monitor Test or COBAS TaqMan HBV Test; Roche Diagnostics, Indianapolis, IN, USA). HBV genotype The major genotypes of HBV were determined using an enzyme-linked immunosorbent assay (ELISA; Institute of Immunology, Tokyo, Japan) or a PCR-invader assay (BML, Tokyo, Japan) according to the methods described by Usuda et al. [16] or Tadokoro et al. [17]. Statistical analysis Differences between groups were examined for statistical significance using the v 2 or Fisher s exact test and Mann Whitney U-test where appropriate. Independent predictive factors associated with response to IFN treatment were determined using multivariate multiple logistic regression. The following 14 potential predictors of response to IFN treatment were assessed in this study: age, sex, pretreatment with IFN, duration of IFN treatment, severity of liver disease (CH or liver cirrhosis), HBV genotype, and levels of aspartate transaminase (AST), ALT, bilirubin, albumin, platelets, a fetoprotein (AFP), HBeAg, and HBV DNA. All factors found to be at least marginally associated with response to IFN therapy (P \.1) were entered into the multivariate multiple logistic regression analysis. The above calculations were performed using the Windows SPSS software package version J (SPSS, Chicago, IL, USA). The odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to assess the relative risk. Independent risk factors predicting the achievement of HBsAg seroclearance were studied using stepwise Cox regression analysis. Potential factors predicting the achievement of HBsAg seroclearance assessed here were the above 14 variables, each transformed into categorical data consisting of two simple ordinal numbers for univariate and multivariate analyses. All factors found to be at least marginally associated with HBsAg seroclearance (P \.1) were tested in the multivariate Cox proportional hazard model. A Kaplan Meier estimate was performed using the SPSS software, and P values were calculated using the Cox- Mantel log-rank test. A two-tailed P value of \.5 was considered statistically significant. Results Study population Twenty-four (4%), 37 (6%), 54 (82%), 1 (.2%), 1 (.2%), and 1 (.2%) patients were infected with HBV genotypes A, B, C, D, H, and B? C, respectively. Genotype could not be measured in the remaining 47 patients. The baseline characteristics of the patients are shown in Table 1. Although few patients had genotypes A and B, the distribution of HBV genotype was similar to that in patients with CHB who had received care in our hospital, with a follow-up period of more than 2 years [18]. Twenty-two of 24 patients with genotype A, 14 of 37 with

4 J Gastroenterol (12) 47: genotype B, 342 of 54 with genotype C, 1 of 1 with genotype H, and 34 of 47 with unknown genotype were HBeAg-positive at the commencement of treatment. While we were able to measure HBV DNA levels in 254 patients at the commencement of IFN therapy, levels in the remaining 361 could not be measured owing to a lack of commercial kits before the bdna assay was available. The numbers of patients receiving other additional therapies after the completion of IFN therapy were 111 (HBeAgpositive/-negative, 9/21), 92 (67/25), 34 (25/9), and 61 (39/22) at the 1-, 3-, 5-, and 1-year time points, respectively. Response to interferon therapy in all patients The IFN response rates in all patients were 21% (15/497), 18% (86/491), 21% (9/428), 23% (82/359), and 25% (59/ 235) at 6 months and 1, 3, 5, and 1 years, respectively, after completion of the IFN therapy (Fig. 1). In patients with genotype A, the response rate was highest at 6 months post-treatment and gradually decreased at subsequent time points from 1 to 1 years thence. In patients with genotype B, response rates were over % at all time points except for 6 months post-treatment, whereas rates in patients with genotype C were under 25% at all time points (Fig. 2a). Evaluation of efficacy of IFN in relation to clinical factors in all patients The data of all patients were subjected to univariate analyses to determine the clinical factors contributing to the efficacy of IFN at each time point. We then investigated the significance of response to IFN therapy using multivariate logistic regression analysis. Interferon response rate (%) ALL HBeAg-positive HBeAg-negative 6 months 1 year 3 years 5 years 1 years 497 (37:127) 491 (367:124) 428 (35:) 359 (247:112) 235 (132:13) Patients observed (n) Fig. 1 Interferon response rates of all patients and hepatitis B e antigen (HBeAg)-positive and HBeAg-negative patients at 6 months and 1, 3, 5, and 1 years Multivariate analyses including the variables noted above revealed several parameters that independently influenced the outcome of IFN therapy; namely, at 6 months: age (P =.13), HBV DNA level (P =.19), and duration of treatment (P =.34); at 1 year: HBV DNA level (P \.1) and age (P =.1); at 3 years: duration of treatment (P \.1), age (P =.13) and albumin level (P =.13); at 5 years: albumin level (P =.4), sex (P =.5), and pretreatment with IFN (P =.39); and at 1 years: HBeAg (P \.1) (Table 2). Response to interferon therapy and evaluation of efficacy of IFN in relation to clinical factors in HBeAg-positive patients Response rates in baseline HBeAg-positive patients were 21% (76/37), 16% (6/367), % (61/35), 21% (53/247), and 18% (24/132) at 6 months and 1, 3, 5, and 1 years, respectively (Fig. 1). In patients with genotype A, the response rate was highest at 6 months post-treatment and the rate was roughly equivalent to the 6 months posttreatment rate at subsequent time points from 1 to 1 years. Response rates in patients with genotype B in particular were above % at all time points except at 6 months, although few patients had genotype B. On the other hand, response rates in patients with genotype C were under % at all time points (Fig. 2a). In addition, multivariate analyses in HBeAg-positive patients also revealed several parameters that independently influenced the outcome of IFN therapy at 6 months: duration of treatment (P =.1) and age (P =.14); at 1 year: age (P =.11) and HBV DNA level (P =.27); at 3 years: sex (P =.8), duration of treatment (P =.19), age (P =.), pretreatment with IFN (P =.29), and albumin level (P =.43); at 5 years: sex (P =.2) and pretreatment with IFN (P =.5); and at 1 years, genotype (P =.19) and AST (P =.35) (Table 3). Response to interferon therapy and evaluation of efficacy of IFN in relation to clinical factors in HBeAg-negative patients Response rates in baseline HBeAg-negative patients were 23% (29/127), 21% (26/124), 24% (29/), 26% (29/112), and 34% (35/13) at 6 months and 1, 3, 5, and 1 years, respectively (Fig. 1). Rates in patients with genotype C were gradually increased at subsequent time points, whereas those in patients with genotype B remained under 3% at all time points (Fig. 2b). In addition, univariate and multivariate analyses in HBeAg-negative patients revealed that duration of treatment (C1 year) independently influenced the outcome of

5 818 J Gastroenterol (12) 47: A Genotype A Genotype B B Genotype A Genotype B 5 Genotype C 1 Genotype C Interferon response rate (%) months 1 year 3 years 5 years 1 years Interferon response rate (%) months 1 year 3 years 5 years 1 years 21:29:1 22:27:2 15::363 12:19:37 5:14:8 Patients observed (n) 19:14:32 :12:35 13:1:259 11:7:212 5:3:1 Patients observed (n) C Genotype A Genotype B Interferon response rate (%) Genotype C 6 months 1 year 3 years 5 years 1 years 2:15:99 2:15:97 2:1:14 1:12:95 :11:88 Patients observed (n) Fig. 2 Interferon response rates of patients with genotypes A, B, and C at 6 months and 1, 3, 5, and 1 years. a All patients, b HBeAg-positive patients, c HBeAg-negative patients IFN therapy at 6 months, and at 1 and 3 years. No parameters independently influenced the outcome of IFN therapy at 5 or 1 years. Evaluation of efficacy of IFN in relation to HBs antigen seroclearance The HBsAg seroclearance rate in this study was obtained from patients who received IFN therapy alone; 69 of 615 patients (11%) achieved seroclearance of HBsAg. The cumulative HBsAg seroclearance rates in all patients from the commencement date of IFN therapy were 6.5% at 5 years, 15% at 1 years, 35% at 15 years, and 44% at years (Kaplan Meier method; Fig. 3a). No patients experienced the reappearance of HBsAg after seroclearance. Five factors found to be associated with achievement of HBsAg seroclearance on univariate analysis were: male sex (P =.2), age C3 years (P =.11), genotype A (P =.38), HBeAg-negativity (P =.45), and bilirubin B1. mg/dl (P =.64). On multivariate analysis, independent factors predicting the achievement of HBsAg seroclearance were: age C3 years, genotype A, and male sex (Table 4). The cumulative HBsAg seroclearance rate for genotype A patients was significantly higher than the rate for those with genotypes B or C (P =.116) (Fig. 3b). Relationship between the response to IFN and the development of hepatocellular carcinoma Twenty-nine patients developed hepatocellular carcinoma (HCC) during the observation period, excluding 17 patients who received other additional therapies after the completion of IFN therapy and developed HCC thereafter. IFN response rates in the 29 patients who developed HCC were 5% (1/22), 5% (1/), 1% (2/), 13% (2/15), and 13% (2/16), respectively, at 6 months and 1, 3, 5, and 1 years after the completion of IFN. No patient developed HCC after HBsAg seroclearance.

6 J Gastroenterol (12) 47: Table 2 Factors associated with response to interferon therapy for all patients at 6 months and 1, 3, 5, and 1 years Parameter Univariate analysis Multivariate analysis OR (95% CI) P OR (95% CI) P 6 Months after completion of IFN therapy (n = 229) Duration of treatment (C1 year) 2.68 ( ) \ ( ).34 HBV DNA level (B7. log copies/ml) ( ) ( ).19 Age (\3 years) ( ).13 1 year after completion of IFN therapy (n = 231) Duration of treatment (C1 year) ( ) \.1 HBV DNA level (B7. log copies/ml) ( ) ( ) \.1 Age (\35 years) ( ) ( ).1 3 years after completion of IFN therapy (n = 397) Duration of treatment (C1 year) 2.41 ( ) \ ( ) \.1 Age (\3 years) 2.7 ( ) ( ).13 Albumin (C3.9 g/dl) ( ) ( ).13 Genotype (non-c) ( ).41 5 years after completion of IFN therapy (n = 356) Albumin (C3.9 g/dl) ( ) ( ).4 Pretreatment with IFN (positive) 1.77 ( ) ( ).39 Sex (female) ( ).5 Duration of treatment (C1 year).8 1 years after completion of IFN therapy (n = 234) HBeAg (negative) ( ) ( ).9 ALT (C1 IU/L) ( ).36 Pretreatment with IFN (positive).58 ALT alanine transaminase, IFN interferon, HBV hepatitis B virus, HBeAg hepatitis B e antigen, CI confidence interval, OR odds ratio, n number submitted to multivariate analysis, including all factors found to be associated with response to IFN therapy Discussion Although IFN has been reported to exert beneficial effects in CHB patients, the response rate is not high. A metaanalysis published in 1993 reviewed 15 randomized controlled studies involving 837 adult patients who received IFN-a for 4 6 months, and elimination of HBeAg occurred in 33% of the treated patients [8]. In previous studies, we found the response rates among HBeAg-positive patients at 6 months after the completion of therapy to be and 31% for 6 months and 1 year of IFN therapy, respectively [13, 15]. Although a recent meta-analysis reported that IFN increased the incidence of HBeAg and HBsAg seroclearance after long-term follow up of 3 7 years [12], the factors that influenced the clinical outcome were unclear. In Japan, from 1988, 4-week IFN treatment was reimbursed by the healthcare system, and since 2, 24-week IFN treatment has been conducted. In the present study, these two regimens were the major ones, and other regimens were used in clinical studies at our hospital (including previously reported studies [14, 15]). Although the durations of treatment differed, we analyzed the factors associated with long-term response to IFN therapy, including the factor of duration of treatment. In the present study, response rates were low among HBeAg-positive patients and relatively high among HBeAg-negative patients at all time points. Approximately % of the HBeAg-positive patients had sustained a response at 6 months to 1 years of follow up. Long-term follow-up studies after a four- to six-month course of IFN therapy in HBeAg-positive patients in European and Taiwanese studies showed higher (33 75%) response rates (HBeAg loss) than our study [7, 19, ]. The difference in response rates between our present study and previous studies in other countries may be due to differences in ethnicity or HBV genotype (mainly genotype C in Japan). Moreover, the low IFN response rates at 1, 3, 5, and 1 years in the HBeAg-positive patients in our study were likely due to the change in treatments (IFN or nucleoside/ nucleotide analogues). On the other hand, the response rates of HBeAg-negative patients in the present study were about % at 6 months and gradually increased thereafter. The sustained response rate in HBeAg-negative patients was usually \3% in European studies [21 23]. The response

7 8 J Gastroenterol (12) 47: Table 3 Factors associated with response to interferon therapy for HBeAg-positive patients at 6 months and 1, 3, 5, and 1 years Parameter Univariate analysis Multivariate analysis OR (95% CI) P OR (95% CI) P 6 months after completion of IFN therapy (n = 279) Duration of treatment (C1 year) ( ) ( ).1 Age (\35 years) ( ) ( ).14 1 year after completion of IFN therapy (n = 172) Duration of treatment (C1 year) ( ).2 HBV DNA level (B7. log copies/ml) 3.59 ( ) ( ).27 Age (\35 years) ( ) ( ).11 3 years after completion of IFN therapy (n = 283) Age (\35 years) 2.41 ( ) ( ). Duration of treatment (C1 year) 2.55 ( ) ( ).19 Pretreatment with IFN (positive) 2.54 ( ) ( ).29 Albumin (C3.9 g/dl) (1. 3.8).43 Sex (female) ( ).8 5 years after completion of IFN therapy (n = 247) Sex (female) ( ) ( ).2 Pretreatment with IFN (positive) 2.46 ( ) ( ).5 1 years after completion of IFN therapy (n = 122) Genotype (non-c) ( ) ( ).19 AST (C1 IU/L) ( ).35 AST aspartate transaminase, IFN interferon, HBV hepatitis B virus, HBeAg hepatitis B e antigen, CI confidence interval, OR odds ratio, n number submitted to multivariate analysis, including all factors found to be associated with response to IFN therapy rates of HBeAg-negative patients in our present study and the studies in other countries [21 23] were similar. Few reports have identified the factors associated with long-term virological response to IFN therapy. In our present study, HBeAg-negativity was the most important factor for predicting a long-term response (1 years). While the HBV DNA level was important for predicting the response at 6 months and 1 year for all patients and the response at 1 year for HBeAg-positive patients, other factors (age, sex, albumin level, AST, IFN pretreatment, and duration of treatment) were found to be important at some time points for all patients and for HBeAg-positive patients. The HBV DNA level may not have been associated with long-term response to IFN therapy because the follow-up period (median 5.7 years) in patients with an HBV DNA level measurable with commercial kits was significantly shorter than that in the other patients (median 11.2 years; P \.1). Previous studies have reported that high ALT levels, low HBV DNA level, female sex, and elevated liver activity and level of fibrosis on liver biopsy were major pretreatment factors correlated with a response to IFN [8 11, 24]. However, in these studies the follow-up times for judging the response were short (typically 6 months to 1 year). Our present study has clarified that HBeAg, HBV DNA level, age, sex, IFN pretreatment, duration of treatment, and levels of albumin and AST are important factors in the long-term response to IFN. Further, non-c genotype was an important factor for long-term response in HBeAg-positive patients. Kao et al. [25] and Lin et al. [] reported that HBV genotype B was associated with a higher response rate to IFN-a therapy than genotype C among CHB patients positive for HBeAg. Similarly, response rates among HBeAg-positive patients with genotype B in the present study were also higher than the response rates in those with genotype C in terms of long-term response (Fig. 2b). The long-term response rate among HBeAg-negative patients was relatively higher than that in HBeAg-positive patients. Previous reports have shown that response rates to a 6- to 12-month course of IFN-a in HBeAg-negative CHB patients range from 1 to 47% (average 24%) [26 29]. In addition, our previous report showed that 9 of 12 (75%) patients who received IFN-b twice per week for 24 weeks responded to the therapy [14]. However, the follow-up periods of these studies were short, and the long-term efficacy has not been clarified. While the efficacy of IFN in HBeAg-negative patients was high in the present study, the factors that might be useful in predicting a sustained response were less well-defined than those in HBeAgpositive patients, as previously reported [5]. A meta-analysis of IFN therapy published in 1 reviewed 6 clinical controlled studies including 828 patients who received IFN [12]. The duration of follow-up

8 J Gastroenterol (12) 47: evelopmen nt rate of HBs antigen n loss (%) de Cumulative f HBs antig gen loss (% %) ent rate of developm Cu umulative (%) A Follow-up period (years) B Genotype A Genotype C P =.116 Genotype B Follow-up period (years) Fig. 3 Cumulative clearance of hepatitis B surface (HBs) antigen in patients treated with interferon (Kaplan Meier method). a All patients, b patients stratified by genotypes A, B, and C Table 4 Factors associated with HBsAg seroclearance by interferon therapy, determined by multivariate analysis Parameter Category Hazard ratio 95% CI P Age \3 years 1.2 C3 years Genotype A 1.4 B C Sex Female 1.5 Male HBsAg hepatitis B surface antigen, CI confidence interval ranged from 35.8 months to 7 years, and HBsAg seroclearance occurred in 9.5% (79/828). In the present study, we observed HBsAg seroclearance in 69 of 615 (11%) patients, with a median follow-up duration of 8.1 years. However, few reports have investigated factors predicting the achievement of HBsAg seroclearance. In our study, important factors for achieving HBsAg seroclearance were age C3 years, genotype A, and male sex. Patients with genotype A had primarily been infected during adulthood via sexual contact, and the average duration of infection was relatively short. In contrast, most Japanese carriers are infected perinatally and possess HBV genotype C, and therefore the efficacy of IFN therapy for patients with genotype C may be low. Male sex was also an important factor in determining potential to achieve HBsAg seroclearance, although female sex was an important factor in determining long-term response to IFN therapy. In our previous study of HBsAg seroclearance (mainly spontaneous seroclearance), we found that response rates were low among females (19%; 45/231) [3]. These present and previous findings indicate that male patients tended to achieve HBsAg seroclearance more frequently than females, although the reason is unclear. We previously reported that Kaplan Meier analysis in 486 patients who received lamivudine therapy for 5 and 1 years showed an estimated loss of HBsAg in 3 and 13% of the patients, respectively, [31]. The cumulative clearance rates of HBsAg, also determined by Kaplan Meier analysis, in patients treated with IFN were higher than those in the patients treated with lamivudine, albeit that there were differences in the baseline characteristics of the patients at the commencement of the respective therapies. The effects of IFN therapy in modulating the host immune response might induce HBsAg clearance. In conclusion, we investigated the long-term efficacy of IFN therapy in Japanese CHB patients. Response rates were low among HBeAg-positive patients and relatively high among HBeAg-negative patients at all time points examined. HBeAg-negative status, HBV DNA level, age, sex, pretreatment with IFN, duration of treatment, and levels of albumin and AST were important factors in predicting long-term response for all patients and for HBeAgpositive patients. Age, genotype, and sex were important factors in predicting ability to achieve HBsAg seroclearance. Further studies exploring the efficacy of therapy over a longer duration may be necessary to confirm these findings and establish true response rates to IFN therapy, including treatment with pegylated IFN. Acknowledgments This study was supported in part by a Grant-inaid from the Ministry of Health, Labor and Welfare of Japan. These authors disclose the following: Dr. Kumada reports having received investigator, lecture, and consulting fees from Dainippon Sumitomo Pharma Co., MSD KK, and Toray Co. Dr. Ikeda reports having received investigator, lecture, and consulting fees from Dainippon Sumitomo Pharma Co. No other potential conflicts of interest relevant to this article were reported.

9 822 J Gastroenterol (12) 47: References 1. Beasley RP, Hwang LW, Lin CC, Chien CS. Hepatocellular carcinoma and hepatitis B virus. A prospective study of 22,77 men in Taiwan. Lancet. 1981;2: Dienstag JL, Perrillo RP, Schiff ER, Bartholomew M, Vicary C, Rubin M. A preliminary trial of lamivudine for chronic hepatitis B infection. N Engl J Med. 1995;333: Marcellin P, Chang TT, Lim SG, Tong MJ, Sievert W, Shiffman ML, et al. Adefovir dipivoxil for the treatment of hepatitis B e antigen-positive chronic hepatitis B. N Engl J Med. 3; 348: Chang TT, Gish RG, de Man R, Gadano A, Sollano J, Chao YC, et al. A comparison of entecavir and lamivudine for HBeAgpositive chronic hepatitis B. N Engl J Med. 6;354: Lok ASF, Heathcote EJ, Hoofnagel JH. Management of hepatitis B: summary of a workshop. Gastroenterology. 1;1: Liaw YF, Sung JJ, Chow WC, Farrell G, Lee CZ, Yuen H, et al. Lamivudine for patients with chronic hepatitis B and advanced liver disease. N Engl J Med. 4;351: van Zonneveld M, Honkoop P, Hansen BE, Niesters HG, Darwish Murad S, de Man RA, et al. Long-term follow-up of alphainterferon treatment of patients with chronic hepatitis B. Hepatology. 4;39: Wong DHK, Cheung AM, O Rourke K, Naylor CD, Detsky AS, Heathcote J. Effect of alpha-interferon treatment in patients with hepatitis B e antigen-positive chronic hepatitis B. Ann Intern Med. 1993;119: Brook MG, Karayiannis P, Thomas HC. Which patients with chronic hepatitis B virus infection will respond to alpha-interferon therapy? Hepatology. 1989;1: Perrillo RP, Schiff ER, Davis GL, Bodenheimer HC Jr, Lindsay K, Payne J, et al. A randomized, controlled trial of interferon alfa- 2b alone and after prednisolone withdrawal for the treatment of chronic hepatitis B. N Engl J Med. 199;323: Perrillo RP, Lai CL, Liaw YF, Dienstag JL, Schiff ER, Schalm SW, et al. Predictors of HBeAg loss after lamivudine treatment for chronic hepatitis B. Hepatology. 2;36: Yang YF, Zhao W, Xia HM, Zhong YD, Huang P, Wen J. Longterm efficacy of interferon alpha therapy on hepatitis B viral replication in patients with chronic hepatitis B: a meta-analysis. Antiviral Res. 1;85: Suzuki F, Arase Y, Akuta N, Tsubota A, Suzuki Y, Sezaki H, et al. Efficacy of 6-month interferon therapy in chronic hepatitis B virus infection in Japan. J Gastroenterol. 4;39: Arase Y, Chayama K, Tsubota A, Murashima N, Suzuki Y, Koida I, et al. A randomized, double-blind, controlled trial of natural interferon-a therapy for e-antigen-negative chronic hepatitis B patients with abnormal transaminase levels. J Gastroenterol. 1996;31: Arase Y, Tsubota A, Saitoh S, Suzuki Y, Kobayashi M, Suzuki F, et al. Randomized, controlled trial of natural interferon-a therapy for e-antigen-positive chronic hepatitis B patients. Hepatol Res. 2;23: Usuda S, Okamoto H, Imawari H, Baba K, Tsuda F, Miyakawa Y, et al. Serological detection of hepatitis B virus genotypes by ELISA with monoclonal antibodies to type-specific epitopes in pres2-region product. J Virol Method. 1999;8: Tadokoro K, Kobayashi M, Yamaguchi T, Suzuki F, Miyauchi S, Egashira T, et al. Classification of hepatitis B virus genotypes by the PCR-Invader method with genotype-specific probes. J Virol Method. 6;138: Kobayashi M, Arase Y, Ikeda K, Tsubota A, Suzuki Y, Saitoh S, et al. Clinical characteristics of patients infected with hepatitis B virus genotypes A, B and C. J Gastroenterol. 2;37: Niderau C, Heintges T, Lange S, Goldmann G, Niederau CM, Mohr L, et al. Long-term follow-up of HBeAg-positive patients treated with interferon alfa for chronic hepatitis B. N Engl J M. 1996;334: Lin SM, Yu KL, Lee CM, Chien RN, Sheen IS, Chu CM, et al. Interferon therapy in HBeAg positive chronic hepatitis reduces progression to cirrhosis and hepatocellular carcinoma. J Hepatol. 7;46: Papatheodoridis GV, Manesis E, Hadziyannis SJ. The long-term outcome of interferon-a treated and untreated patients with HBeAg-negative chronic hepatitis B. J Hepatol. 1;34: Brunetto MR, Oliveri F, Coco B, Leandro G, Colombatto P, Gorin JM, et al. Outcome of anti-hbe positive chronic hepatitis B in alpha-interferon treated and untreated patients: a long term cohort study. J Hepatol. 2;36: Lampertico P, Ninno ED, Vigano M, Romeo R, Donato MF, Sablon E, et al. Long-term suppression of hepatitis B e antigennegative chronic hepatitis B by 24-month interferon therapy. Hepatology. 3;37: Lau DTY, Everhart J, Kleiner DE, Park Y, Vergalla J, Schmid P, et al. Long-term follow-up of patients with chronic hepatitis B treated with interferon alfa. Gastroenterology. 1997;113: Kao JH, Wu NH, Chen PJ, Lai MY, Chen DS. Hepatitis B genotypes and the response to interferon therapy. J Hepatol. ;33: Hadziyannis S, Bramou T, Makris A, Moussoulis G, Zignego L, Papaioannou C. Interferon alfa-2b treatment of HBeAg negative/ serum HBV DNA positive chronic active hepatitis B. J Hepatol. 199;11:S Pastore G, Santantonio T, Milella A, Monno L, Mariano N, Moschetta R, et al. Anti-HBe-positive chronic hepatitis B with HBV-DNA in serum: response to a 6-month course of lymphoblastoid interferon. J Hepatol. 1992;: Fattovich G, Farci P, Rugge M, Brollo G, Mandas A, Pontisso P, et al. A randomized, controlled trial of lymphoblastoid interferonalfa in patients with chronic hepatitis B lacking HBeAg. Hepatology. 1992;15: Lampertico P, Del Ninno E, Manzin A, Donato MF, Rumi MG, Lunghi G, et al. A randomized, controlled trial of a 24-month course of interferon alfa 2b in patients with chronic hepatitis B who had hepatitis B virus DNA without hepatitis B e antigen in serum. Hepatology. 1997;26: Arase Y, Ikeda K, Suzuki F, Suzuki Y, Saitoh S, Kobayashi M, et al. Long-term outcome after hepatitis B surface antigen seroclearance in patients with chronic hepatitis B. Am J Med. 6;119:71 e9 e Kobayashi M, Suzuki F, Akuta N, Hosaka T, Sezaki H, Yatsuji H, et al. Loss of hepatitis B surface antigen from the serum of patients with chronic hepatitis treated with lamivudine. J Med Virol. 7;79: